- Google launched an internal AI code assistant called Goose to accelerate developer productivity.
- With a 28,000 token context, Goose provides code edits and answers on Google’s technologies.
- The AI tool aims to optimize coding speed across the tech giant if effective.
Google seeks any edge to accelerate development, now turning to AI itself. Dubbed “Goose”, the large language model descendant helps engineers write code.
Supporting the drive toward efficiency
Goose brings 25 years of Google expertise to bear, answering developer questions on internal technologies and even editing code.
The launch supports Google’s drive toward efficiency and streamlined operations, hinted at in recent earnings.
Microsoft deployed a similar internal tool in Copilot. But Goose’s wide context window of 28,000 tokens aids in processing long code.
It can also utilize the twin strengths of Google Brain and DeepMind after their merger.
A significant gains in coding speed
With Goose recommending code optimizations and an AI bot assistant named Duckie to resolve issues, Google aims to bring “AI to every stage” of building products.
If it works, Goose could hatch significant gains in coding speed across the tech giant.